We traded one form of division for another
Humanity built social media to close distances between people, only to watch it widen the distances within them. Now, a new technology arrives carrying the same promise — artificial intelligence as a bridge across the fractures that algorithms once deepened. Yet the history of tools reminds us that the power to unite is inseparable from the power to control, and the question of who holds that power may matter more than the technology itself.
- Social media's algorithms quietly rewarded outrage over understanding, sorting societies into sealed information worlds where the other side became not just wrong but threatening.
- AI is now being proposed as the antidote — a system capable of surfacing shared facts, translating between divided groups, and steering public discourse toward common ground rather than conflict.
- The danger is that AI's very strength — its ability to invisibly shape how people perceive reality — makes it a more potent instrument of control than social media ever was, with fewer visible seams to pull at.
- A handful of companies already hold the levers of this new system, raising the prospect that we may be trading visible algorithmic division for something quieter, deeper, and harder to resist.
- The unresolved tension sits here: the promise of unity is genuine, but without deliberate choices about governance and accountability, AI may simply replace one hierarchy of influence with a more sophisticated and invisible one.
We built social media to connect us, and for a time it seemed possible. Then the algorithms discovered that anger kept people scrolling, and the feeds filled accordingly. By the time the damage was understood, it was structural — separate information ecosystems, each convinced the other was not just mistaken but dangerous.
Now AI is being offered as the remedy. The logic has appeal: a system trained toward common ground rather than conflict, one that surfaces shared facts instead of inflaming differences, that helps people see across the walls that social media built. The vision is coherent. The problem is what it requires.
The same quality that makes AI promising as a unifying force — its capacity to shape how we understand the world — is precisely what should give us pause. Social media divided us because a small number of companies controlled what we saw. The proposed solution is another system controlled by a small number of companies, except this one operates more quietly, more deeply, and with less visible machinery to inspect or challenge.
Where social media at least made its divisions legible — you could sense the algorithm was feeding you something different than your neighbor — AI can shape consensus without the shaping being noticed. It can present itself as a neutral arbiter while making consequential choices about what counts as truth and what problems deserve attention.
The real question is not whether AI could theoretically unite us. It probably could. The question is whether we are prepared to reckon with the cost: concentrated power, invisible influence, and diminished control over how our picture of the world is assembled. We recognized social media as a problem of power and incentives. AI presents the same problem in a more refined form.
What determines the outcome is not the technology but the choices made about who governs it and toward what ends. Without those choices, we risk trading one division for another — between those who understand how the system works and those quietly shaped by it, between those who hold the tool and those the tool is held over.
We built social media to connect us, and for a moment it seemed like it might work. Instead, the algorithms learned to reward outrage. They learned that people stay longer when they're angry at each other, and so the feeds filled with content designed to deepen every fracture in the culture. By the time we understood what was happening, the damage was already structural. We had sorted ourselves into separate information ecosystems, each one convinced the other side was not just wrong but dangerous.
Now there is talk of AI as the antidote. The logic is straightforward enough: artificial intelligence could be trained to find common ground, to surface the facts both sides might agree on, to build bridges where algorithms once built walls. Instead of feeding us content designed to inflame, AI could serve us information shaped toward understanding. It could translate between groups. It could help us see each other again.
But here is where the problem begins. The very thing that makes AI promising as a unifying force—its power to shape how we see the world—is also the thing that should make us nervous. Social media divided us because a handful of companies controlled the levers. They decided what we saw and in what order. We are now talking about replacing that system with another one controlled by a handful of companies, except this time the tool is more sophisticated, more invisible, and harder to resist.
The concentration of AI power in a few hands creates a new kind of vulnerability. Where social media at least made its divisions visible—you could see the algorithm was showing you different content than your neighbor—AI systems work more quietly. They can shape consensus without anyone noticing the shaping is happening. They can present themselves as neutral arbiters when they are actually making choices about what counts as truth, what deserves attention, what problems matter most.
There is a genuine appeal to the idea of AI as a unifying technology. We are tired of being divided. We want to believe there is a tool that could help us find our way back to shared reality. But the history of technology suggests that the same tool that solves one problem often creates another. Social media promised to democratize information and connect the world. It did both those things. It also fragmented our sense of what is real and turned engagement into a business model that rewards division.
The question is not whether AI could theoretically unite us. It probably could. The question is whether we are willing to accept the cost of that unity—the concentration of power, the invisibility of influence, the loss of control over how our understanding of the world is shaped. We solved the problem of social media division by recognizing it as a problem of power and incentives. We need to recognize that AI presents the same problem in a more sophisticated form.
What matters now is not the technology itself but the choices we make about who controls it and how. Without those choices, we will have simply traded one form of division for another—a division between those who understand how the system works and those who do not, between those who control the AI and those who are shaped by it. The promise of unity is real. So is the risk.
Citações Notáveis
AI systems work more quietly than social media—they can shape consensus without anyone noticing the shaping is happening— Analysis of AI's structural advantages over social media
A Conversa do Hearth Outra perspectiva sobre a história
You're saying AI could actually bring people together in a way social media couldn't?
In theory, yes. AI could find the facts both sides agree on, present information in ways that build understanding instead of triggering defensiveness. It could work at scale in ways human dialogue can't.
But you're worried about something else.
I'm worried about who controls it. Social media divided us, but at least we could see the division happening. AI works invisibly. It shapes what we think is true without us noticing the shaping.
So it's the same problem—power concentrated in a few hands—just harder to see?
Exactly. We learned that lesson with social media. The technology itself wasn't the problem. The problem was that a handful of companies decided what billions of people saw. Now we're about to hand them a more powerful tool.
What would it look like if we got this right?
Transparency about how the system works. Real limits on who can build and control these systems. A genuine commitment to keeping power distributed instead of concentrated. But I'm not sure we're having that conversation yet.
And if we don't?
We'll have solved the problem of division by creating a new problem—one where most people don't even realize they're being shaped.